Software Design Of Material Composition Detection System Based On Micro Spectrometer

The micro spectrometer has the advantages of small size,light weight,convenient carrying and low cost,and is an important branch of the spectrometer.The traditional software used with the spectrometer consists of two parts,the instrument control software and the chemometrics software.The separation of the two parts brings problems such as inconvenience to the user and inability to perform online detection.Based on this,it is significance to study the software of material composition detection system which is matched with micro spectrometer.Based on the analysis of micro spectrometer software design and spectral pretreatment methods,this paper designs a set of software for material composition detection system based on micro spectrometer.This paper solves the problem that the separation of the micro spectrometer instrument control software and stoichiometry software,and makes it impossible to detect the components online.The main work and research results of this paper are listed as follows:(1)Design of micro spectrometer calibration and denoising algorithm: first,by analyzing the micro spectrometer hardware system and optical path system,the calibration method of spectrometer based on moving least squares method is designed for the problem of traditional spectrometer calibration complex and low precision.Then,for the problem of high frequency noise in the micro spectrometer,the average-wavelet denoising method is designed to eliminate the high frequency noise of the spectral image while retaining the spectral features.Finally,in order to meet the needs of FWHM(full wave half maximum)measurement in spectral analysis,a measurement method with Gaussian fitting and peak-seeking algorithm is designed.(2)Software design of the substance composition detection system: first,the thesis analyzes the user requirements and designs the overall framework of the software system.Then,the user interaction interface,serial communication,image display,data storage and readined functions are designed on the WPF(Windows Presentation Foundation)platform.Finally,in order to solve the problem of inconvenient use caused by the separation of traditional instrument control and stoichiometric software,the calibration and usage procedures of linear and nonlinearmaterial models are designed.The material detection calibration model based on BP neural network is designed,and the thesis realizes the online detection of substance.(3)System function test and result analysis: first,the calibration accuracy and spectral image denoising quality of the micro-spectrometer's material composition detection system are tested and evaluated,and then the laser light source is used to test the stability of the micro spectral detection system to measure the full width at half maximum.Next,the nitrate nitrogen sample solution is configured to model the experimental sample.Finally,the performance of the system is verified by online detection of the nitrate nitrogen content in the test sample,and it is verified that the software of the material composition detection system designed in this thesis has high detection accuracy.